Air pollution has been linked to a number of health impacts and has been studied in a variety of contexts using a variety of studies and methodologies. This thesis is made up of a collection of papers that cover a wide range of research subjects and illustrate different study analysis and design methodologies. Multiple imputation (MI) techniques were used to deal with the missing data, where missForest had the lowest imputation error among the other imputation approaches. Time series modelling was used to predict Rheumatoid Arthritis (RA) disease activity score (DAS28) using the information of air pollution. This thesis examined the linkage among SO2, NO2, O3 and disease activity scores for patients with RA in Kuwait. The association was investigated using the Granger causality test (using the VECM approach and other time series approaches) (in analysis of static causality) and the Impulse Response Functions (IRFs) analysis (in analysis of dynamic causality). A comprehensive conceptual framework was used in the study, which included a cointegration test, unit root test, and panel VECM. Long-run causation and asymptotic convergence among the variables were determined using the panel VECM. The empirical outcomes show that NO2 and O3 are statistically significant in cases when DAS28 is the dependent variable, in most of the study locations (ASA, FAH, MAN and JAH). The results demonstrate that the lagged error correction term (ECT) coefficients in DAS28 and air pollution emissions are statistically significant. Overall, the main conclusion found in this thesis and according to the cointegration test, the results show that there exists a long run relationship between the emissions of air pollution and the change of DAS28 among RA patients.
Date of Award | 1 Sept 2022 |
---|
Original language | English |
---|
Awarding Institution | - University Of Strathclyde
|
---|
Supervisor | Jiazhu Pan (Supervisor) & Xuerong Mao (Supervisor) |
---|